TOMASZ BIGAJ
THE INDISPENSABILITY ARGUMENT –
A NEW CHANCE FOR EMPIRICISM IN MATHEMATICS?
ABSTRACT. In recent years, the so-called indispensability argument has been
given a lot of attention by philosophers of mathematics. This argument for the
existence of mathematical objects makes use of the fact, neglected in classical
schools of philosophy of mathematics, that mathematics is part of our best
scientific theories, and therefore should receive similar support to these theories.
However, this observation raises the question about the exact nature of the alleged
connection between experience and mathematics (for example: is it possible to
falsify empirically any mathematical theorems?). In my paper I would like to
address this question by considering the explicit assumptions of different versions
of the indispensability argument. My primary claim is that there are at least three
distinct versions of the indispensability argument (and it can be even suggested
that a fourth, separate version should be formulated). I will mainly concentrate
my discussion on this variant of the argument, which suggests the possibility of
empirical confirmation of mathematical theories. A large portion of my paper will
focus on the recent discussion of this topic, starting from the paper by E. Sober,
who in my opinion put reasonable requirements on what is to be counted as an
empirical confirmation of a mathematical theory. I will develop his model into
three separate scenarios of possible empirical confirmation of mathematics. Using
an example of Hilbert space in quantum mechanical description I will show that
the most promising scenario of empirical verification of mathematical theories
has nevertheless untenable consequences. It will be hypothesized that the source
of this untenability lies in a specific role which mathematical theories play in
empirical science, and that what is subject to empirical verification is not the
mathematics used, but the representability assumptions. Further I will undertake
the problem of how to reconcile the alleged empirical verification of mathema-
tics with scientific practice. I will refer to the polemics between P. Maddy and
M. Resnik, pointing out certain ambiguities of their arguments whose source is
partly the failure to distinguish carefully between different senses of the indis-
pensability argument. For that reason typical arguments used in the discussion are
not decisive, yet if we take into account some metalogical properties of applied
mathematics, then the thesis that mathematics has strong links with experience
seems to be highly improbable.
KEY WORDS: applicability of mathematics, confirmation, empiricism, mathe-
matical realism, quantum mechanics
Foundations of Science 8: 173–200, 2003.
© 2003 Kluwer Academic Publishers. Printed in the Netherlands.